Perhaps programmers have their own unit tests, perhaps as the result of doing Test-Driven Development. They run these tests very often; perhaps every few minutes, but at least a few times each day.
In my experience, QA's automated tests tend to go after bigger chunks of functionality and are therefore slower and usually less reliable than programmer's unit tests. Still, there is a lot of value in getting the results of these tests in programmer's hands as soon as possible.
Should QA push for programmers running QA's automated tests? Should QA insist that all of their automated tests pass before checking in? Should the nightly build process include a run of QA's automated tests?
(I wrote nightly build but that only makes sense if you have a nightly build. I suspect that many shops don't have this. I suspect that how QA is done varies very widely, and that awareness of other ways of doing things is very limited.)